Numerical Technologies Random Generator for Excel
NtRand, Numerical Technologies’s Random Number Generator for Microsoft Excel is a software where a family of various random numbers such as standard normal (Gaussian) random numbers in Monte Carlo method is implemented as add-in functions available with Excel. It also generates multivariate correlation random numbers.
NtRand uses pseudo-random numbers by Mersenne Twister algorithm,
featuring longer period and higher order of equidistribution, which is a cynosure in simulation technology.
Multivariate Mote Carlo employs antithetic variant method and quadratic resampling for higher accuracy. For more information, see “Enhancement of Monte Carlo by Moment Matching.”
In the VaR computing for risk management, a larger scale Monte Carlo simulation is often necessary. In this case, the built-in random numbers would not serve well enough in ensuring the accuracy. Some solutions use the physical random numbers.
However, they are less practical because no significant repeatability is observed on any verification test and some special hardware is required. Mersenne Twister is one of the approaches to provide a pragmatic solution in VaR quantification, and widely incorporated in PortfolioBrowser® / CreditBrowser®, risk management system developed by Numerical Technologies Incorporated.
NtRand operates at very high speed with many advanced features. It enables you to build a several tens of dimensions of practical Monte Carlo VaR calculating system or derivatives pricing model with the Excel running on your own computer. That is, almost all smaller-scaled VaR management systems can be configured using NtRand. We recommend NtRand for derivatives pricing, risk management studies, preliminary test before introducing a costly calculating system, or writing your paper.
|Outstanding properties of longer period and higher order of equidistribution for pseudo-random numbers (random period of 219937-1). For further information, visit the Mersenne Twister Home Page hosted by assistant professor Makoto Matsumoto and Takuji Nishimura at Faculty of Science and Technology, Keio University, Japan.|
|Easy generation of the multivariate correlation random numbers essential to the Monte Carlo VaR calculation.|
|High speed operation enhanced by the efficient design in engineering. NtRand operates far faster than RAND function in Excel and easily generates random numbers for more than 10,000-time Monte Carlo.|
|Various moment matching methods available to improve the convergence in Monte Carlo. Enables you to easily use the antithetic variant method and/or quadratic resampling. For more information, see “Enhancement of Monte Carlo by Moment Matching.”|
|Utilities to supplement the improvement of Monte Carlo’s convergence – For uniform/normal random numbers conversion, the Box-Muller method is available in addition to the inverse function method (Moro’s algorithm).|
|Random number generator and related utility functions for widely acknowledged following 17 different probability distributions widely acknowledged.|
(*) Mom. function returns mean, standard deviation, skewness and kurtosis of the distribution at once.
- – Supported
- – Supported in Excel
- – Not supported